The control of human globin gene expression in erythroid cells involves trans-factors (substances active at distant locations in the genome), which have yet to be identified or clearly described. One experimental approach to their identification is to study the effects on globin gene expression of well-described trans-factors from tumor viruses. We have shown that the HTLV-I trans-factor tat-I stimulates both beta- and epsilon-promoters fused to a CAT gene, resulting in roughly 20-fold increase in CAT enzyme activity. In the case of beta-globin, only 185 bp of 5' flanking sequence is required for this effect. Recently, we have shown that stimluation of the epsilon-globin promoter requires at least 114 but not more than 177 bp of 5' flanking sequence. Computer analysis reveals 2 elements in this region homologous to the tat-I response element of HTLV-I; their functional significance is being investigated. Further studies will involve characterization of the tat-I induced trans-activation of globin promoters. In progress are studies of the cis elements involved in tat-I responsiveness. These include fine structure genetic mapping by site directed mutagenesis of suspected response elements. Identification of these cis elements may lead to identification of DNA binding proteins involved in trans-activation of globin genes by tat-I. Our ultimate objective is to identify such cellular proteins that interact with tat-I to trans-activate globin genes. Study of such proteins may clarify the developmental regulation of globin gene expression in human erythroid cells.